the body × the ledger · crossing · EP-036

How to Read a Health Headline Without Being Lied To

"The study says" is the modern oracle — and it is mostly misread. The Body meets the Ledger: five questions that disarm almost any health headline (in whom, compared to what, relative or absolute, who paid, one or many), a real-shaped headline worked through…

Airs Monday 11 January 2027 on the channel. The video appears here the day it is published.

"The study says" is the modern oracle — and it is mostly misread. The Body meets the Ledger: five questions that disarm almost any health headline (in whom, compared to what, relative or absolute, who paid, one or many), a real-shaped headline worked through all five by hand, and the posture that protects you from both the swallowers and the sneerers. Strong claims move slowly. The subjects of the basics named near the end — sleep, physical activity, diet and tobacco — are on the American Heart Association's list of eight measures of cardiovascular health. Media literacy, not medical advice — anything urgent in your own body belongs to emergency care.

Drawn from: plain research-literacy facts — study populations, control groups, observational versus assigned designs, funding interest, and replication — plus worked relative-versus-absolute arithmetic worked through in the episode (a 40% cut from 5 in 1,000 lands at 3 in 1,000). The base-rate machinery builds on this channel's EP-013.

General education, not advice. Sources are named on screen and in the description; where a guideline body is named, its current guidance governs. The disclaimer.

as spoken

The words, in order.

  1. Forty mice became your breakfast news.

    In a laboratory somewhere, forty mice had an interesting week. By Thursday it was a headline about your breakfast, and by Friday someone at your table was repeating it as settled law.

  2. The modern oracle, mostly misread.

    "The study says." That is the modern oracle — consulted like scripture, read like a horoscope. Today the Body meets the Ledger at a crossing: how to read a health claim without being lied to, using five questions you can count on one hand.

  3. The science is fine. The line is long.

    First, respect where it is owed: the method itself is one of the best tools our species has built for finding out what is true. The problem is not the science. The problem is the long telephone line between the laboratory and the headline — and what happens to a sentence traveling it.

  4. In whom — and how well built?

    Question one: in whom? Mice are not men. Cells in a dish are not even mice. And counting heads settles nothing by itself — twelve volunteers over two weeks is thin ice, but a huge, careless study is only a louder rumor, and a small, careful one can outweigh it. Ask how far the study stands from your actual body — and how well the thing was built.

  5. Compared to what?

    The second question is older than science itself: compared to what? Nothing is good or bad alone — only against the alternative. Better than nothing? Better than the old treatment? Better than a sugar pill and a kind nurse? A control group gives the claim something honest to stand against; without one, you are holding half a comparison.

  6. Make the percentage confess its base.

    Question three you have drawn from this well before: relative or absolute? "Cuts risk by forty percent" must be made to confess its starting number. Forty percent of rare is still rare — the base rate is the riverbed under every headline.

  7. Run the five on it.

    Now take one home and work it. The headline reads: a daily handful of nuts cuts heart risk by forty percent. Six words of it are doing all the lifting. Run the questions.

  8. The nuts keep respectable company.

    In whom? Say thirty thousand adults followed for ten years — a serious study, credit where due. Compared to what? To their neighbors who skipped the nuts — but observed, not assigned: nobody was ordered to eat them. And people who choose a daily handful of nuts differ in a dozen quiet ways — they walk more, smoke less, see doctors sooner. The nuts keep respectable company, and the company may be doing the work.

  9. From five in a thousand to three.

    Relative or absolute? If the yearly risk was five in a thousand, a forty percent cut brings it to three in a thousand. Two people per thousand per year — worth having, if the effect is real and the nuts, not their respectable company, deserve the credit. But the honest sentence, "from five in a thousand to three," rarely makes a front page.

  10. True sentences, kindly chosen.

    The fourth question follows the money: who paid? Not an accusation — an accounting. Disclosure is honesty, not guilt, and a nut council can fund honest work. But read sponsored findings the way you read a seller's brochure: every sentence may be true, and every sentence was chosen. Ask what was measured and never mentioned.

  11. One study is one witness.

    And the last question: one study, or the weight of many? A single study is a single witness, and science is a courtroom that likes several. The startling lone result is sometimes a herald, more often an outlier — each careful repetition that holds is what turns a claim into knowledge, and tomorrow's repeat has a long record of shrinking today's miracle.

  12. The headline is priced in clicks.

    Here the Ledger explains the Body's predicament: the incentive runs against you. The dull true sentence does not sell, and the seller knows it. Few outlets print "large study finds modest effect, probably." The headline is priced in clicks, and the currency it spends is you.

  13. Truth compounds. News spikes.

    So hold the honest posture: strong claims move slowly. What is actually known about the body has been boringly true for decades — sleep, movement, plants on the plate, no smoke in the lungs. Truth compounds like savings. News spikes like a fever.

  14. Neither swallow nor sneer.

    But do not fall into the opposite ditch — the sneer that says the studies always change their minds, so nothing is known. That is the same laziness wearing a smirk. The five questions are the narrow path between swallowing everything and spitting on sight.

  15. Five questions before you repeat it.

    Here is your work. The next health headline that reaches you — the week will supply one — run the five aloud before you repeat it to anyone: in whom, compared to what, relative or absolute, who paid, one or many. If it survives all five, share it with its numbers attached. And know the tool's edge: these questions read headlines, not bodies. What is urgent in your own chest belongs to emergency care, not a checklist.

read alongside it

From the shelf.

As an Amazon Associate I earn from qualifying purchases. The commission costs you nothing and bought none of these; the whole shelf is on the reading page, and how that works is on the about page.

  1. How to Lie with Statistics — Darrell Huff

  2. Naked Statistics — Charles Wheelan

  3. How Not to Be Wrong — Jordan Ellenberg

keep drawing

More from The Body and The Ledger.

drink deep

Here is the map.

The Codex of Nine is a free twelve-page primer: the well beneath the tree, then one page for each domain — the question it answers, three truths already drawn, and one thing to do this week.

After that, one note per episode — the video and its sources — twice a week and nothing else. Unsubscribe in one click. Privacy.